Incremental innovation lives on

In 2010, in my blog, Drug­Baron, I de­clared the death of in­cre­men­tal in­no­va­tion and wrote the obit­u­ary.

It didn’t seem a very bold state­ment at the time - in­creas­ing payor pres­sure was be­gin­ning to bite, and it seemed ob­vi­ous that the en­trenched prac­tice of pay­ing a pre­mium for small (often im­per­cep­ti­ble) im­prove­ments was pass­ing. If pay­ors, whether gov­ern­ments, pri­vate in­sur­ers or even pa­tients, were un­will­ing to pay for in­cre­men­tal in­no­va­tion, then the in­escapable logic seemed to be that the in­dus­try would have to stop de­vel­op­ing such prod­ucts, and in­stead seek big­ger im­prove­ments in clin­i­cal per­for­mance.

But three years later, the mar­ket for late-stage prod­ucts that show lit­tle, if any, dif­fer­en­ti­a­tion re­mains as frothy as it has ever been.  As if to prove the point, As­trazeneca in­vested a lit­tle over a bil­lion dol­lars in two such prod­ucts in one month: an­other LABA com­bi­na­tion from Pearl, and an­other pre­scrip­tion fish oil cap­sule from Omthera.

With deals like this on the table, why not de­velop an­other me-too prod­uct against a well-val­i­dated human tar­get?  After all, the tech­ni­cal risk is very low.  So as long as you as are con­fi­dent some­one will ac­quire the prod­uct, it looks like an ex­cel­lent in­vest­ment. In com­mon with Mark Twain, then, re­ports of the death of in­cre­men­tal in­no­va­tion were clearly ex­ag­ger­ated. How then can you square the cir­cle? 

If pay­ors in­creas­ingly refuse to pay for in­cre­men­tal in­no­va­tion (even be­fore 'pay for per­for­mance' re­ally takes off, as it surely will), yet pharma com­pa­nies are pay­ing big bucks for poorly dif­fer­en­ti­ated prod­ucts then some­one has to be mak­ing a costly error. The an­swer is that pharma com­pa­nies (or some of them at least) are using their gar­gan­tuan bal­ance sheets to ab­sorb the pres­sure from pay­ors, rather than trans­mit­ting it to the early stage dis­cov­ery ef­forts (whether in­ter­nally or ex­ter­nally).  While they con­tinue to buy poorly dif­fer­en­ti­ated late-stage prod­ucts, the drive to dis­cover and de­velop re­ally risky as­sets that might de­liver a step change in out­comes re­mains im­po­tent.

The big ques­tion has to be why do they do it? After all, it is a strat­egy that is likely to end badly.  If you fill a pipeline with prod­ucts dis­play­ing lit­tle real in­no­va­tion or im­prove­ment in out­comes against a land­scape where in­creas­ingly no-one will pay for such prod­ucts then you are set­ting your­self up for cat­a­strophic fail­ure.  You can only ig­nore the de­mands of your cus­tomers for so long, even in the Alice in Won­der­land world of phar­ma­ceu­ti­cals.  A big bal­ance sheet only pro­vides so much in­su­la­tion. But that fail­ure is a long time com­ing.  And if you are very big, it could be a decade or more away. 

It seems that there are much more short term con­cerns that de­mand ac­tion - and lead to be­hav­iours that are ul­ti­mately close to sui­ci­dal. Elias Za­houni, the elo­quent R&D chief at Sanofi re­cently hinted at the un­der­ly­ing cause of these con­flict­ing be­hav­iours: large pharma com­pa­nies are prin­ci­pally mo­ti­vated by fear, and ex­pend 90% of their ef­fort on de­fend­ing ex­ist­ing rev­enues rather than on gen­er­at­ing new sources of in­come.  Such de­fence is pru­dent and nec­es­sary, but the bal­ance is wrong, he says - maybe half and half be­tween de­fence and in­no­va­tion would be a health­ier mix, with maybe 10% on the re­ally wild-eyed vi­sion. This de­fen­sive at­ti­tude causes pharma to ac­quire prod­ucts that are less risky (and hence less in­no­v­a­tive) than their cus­tomers are de­mand­ing. 

That is a very dan­ger­ous mis-match - and one that be­comes more se­vere the emp­tier the pipeline is per­ceived to be. For a pub­licly-traded com­pany, the spot­light falls heav­ily on the late-stage pipeline.  And bad news trans­lates quickly into a falling mar­ket cap.  Faced with a choice of two kinds of bad news - a Ph3 clin­i­cal trial fail­ure or the fail­ure of a prod­uct to win a sub­stan­tial mar­ket (prod­ucts that I have pre­vi­ously called dubbed 'busters', as op­posed to 'block­busters'), then the de­fen­sive ex­ec­u­tive will al­ways chose the lat­ter.  After all, mar­ket fail­ure will only be ob­vi­ous nearly a decade down the line.  A phase 3 fail­ure could hit the value of the com­pany to­mor­row.

Far bet­ter, then, to ac­quire a prod­uct with so lit­tle in­no­va­tion that it can hardly fail to be ap­proved - and then sweat about whether pay­ors will ac­tu­ally pay for it in any kind of vol­ume - than it is to ac­quire a gen­uinely in­no­v­a­tive, first-in-class prod­uct and then dis­cover it has feet of clay. And the more your in­no­v­a­tive bets tum­ble (as Lilly as ex­pe­ri­enced re­cently), the greater the pres­sure on the ex­ec­u­tives to flex the bal­ance sheet and ac­quire "safe" bets.

The re­sult will surely be a vis­cous down­wards cir­cle, where ac­qui­si­tions of un­dif­fer­en­ti­ated prod­ucts that can­not fail in the clinic leads to too few sales to jus­tify the pur­chase price.  There has to be a very real risk that some large pharma are al­ready un­com­fort­ably close to this 'death spi­ral'. Avoid­ing this fate isn't just about tak­ing more risk, even if the pub­lic mar­ket in­vestors would tol­er­ate it. 

The in­dus­try has to get smarter at tak­ing these risks.  Right now, too many in­no­v­a­tive first-in-class pro­grammes are fail­ing in Phase 3. The prob­lem is not that real in­no­va­tion is risky - of course it is.  But a way has to be found to fail these pro­grammes much ear­lier.  I have al­ready ex­am­ined the fac­tors that led sev­eral pharma com­pa­nies to progress anti-amy­loid drug can­di­dates into Phase 3 (and in­deed to keep on try­ing) when all the data (even from Phase 2) sug­gested this class of agents was doomed to fail. 

The same story ap­plied to Merck with the CETP in­hibitor anace­trapib, and most re­cently to As­trazeneca with fos­ta­ma­tinib.  To his credit, at least Pas­cal So­riot and his team ac­cepted the fail­ure, took the painful med­i­cine, killed the fos­ta­ma­tinib pro­gramme and moved on.  Anti-amy­loids and CETP in­hibitors, though, con­tinue to con­sume vast re­sources with lit­tle hope of even­tual suc­cess. Tak­ing the risks nec­es­sary to de­liver real ad­vances in pa­tient care are only eco­nomic if the dis­cov­ery and de­vel­op­ment costs can be shrunk - not just 20-30% as has been widely ac­cepted across the in­dus­try but by two-fold or even more.

That kind of cost re­duc­tion can­not be achieved by evolv­ing the old model - a model that in any case grew in a very dif­fer­ent world: a world of 'gold rush eco­nom­ics' where any ap­proved drug au­to­mat­i­cally gar­nered suf­fi­cient sales to jus­tify dis­cov­er­ing a new drug at any price.  Dis­cov­ery was op­ti­mised for suc­cess not pro­duc­tiv­ity. A change in the way we de­sign and in­ter­pret Phase 2a stud­ies, the key gate­keeper in the drug de­vel­op­ment process, lying be­tween rel­a­tively low cost early de­vel­op­ment and the mega-bucks of late stage, is the most press­ing need.  Stud­ies must be able to dis­tin­guish those prod­ucts that will suc­ceed, not just in later de­vel­op­ment but also in the mar­ket­place, from those that are doomed to fail sooner or later.

Yet again, though, the con­ser­v­a­tive mind­set in large com­pa­nies makes such changes dif­fi­cult.  If a study is bor­der­line, or yields a grey re­sult, its bet­ter to keep the pipeline full and delay the bad news for an­other day.  After all, it might work. And its not just large com­pa­nies.  The de­fen­sive mind­set can per­me­ate even the small­est op­er­a­tions if the man­age­ment feel that their jobs, or even their ca­reers, are in peril should the prod­uct they are de­vel­op­ing fail. 

The in­cen­tive to delay fail­ure, until later in de­vel­op­ment, or even into the mar­ket­place, are very strong in­deed - even while re­peat­ing the 'fail early, fail cheap' mantra, ex­ec­u­tives in drug com­pa­nies large and small are, it seems, still be­hav­ing in ex­actly the op­po­site fash­ion. So, in­cre­men­tal in­no­va­tion is alive and well.  Even as it’s starved of oxy­gen by the tight­en­ing purse strings of the even­tual cus­tomers, it is being sus­tained by the stronger im­per­a­tive to avoid clin­i­cal fail­ures, or any kind of im­mi­nent bad news, by pharma com­pa­nies who failed to grasp the need to change to a leaner dis­cov­ery and de­vel­op­ment model a decade ago. 

For those that did start that process ear­lier, and whose late stage pipeline looks health­ier as a re­sult, they need to stick with it - and cut costs even more ag­gres­sively, al­low­ing them to keep up the level of risk they can take to de­liver real in­no­va­tion. Only they will sur­vive.  Any that are caught in the trap of doing low-risk in­cre­men­tal in­no­va­tion when the cus­tomer de­mands a step-change in out­comes, de­fy­ing logic through the sheer power of their bal­ance sheets built up on his­tor­i­cal suc­cesses, will end up com­pletely drained.

For early stage in­vestors, then, there is a real co­nun­drum - should they back poorly dif­fer­en­ti­ated prod­ucts be­cause they offer good re­turns at lower risk as long as pharma keep buy­ing?  That de­pends on how long the largest com­pa­nies con­tinue to use their fi­nan­cial mus­cle to warp re­al­ity.  I was three years too early call­ing the death of in­cre­men­tal in­no­va­tion - and he is not about to re­peat that mis­take.  In­cre­men­tal in­no­va­tion ought to be dead, but it looks like it will be around for a while yet.